Brian's Blog

I've noticed that Nanoblogger 3.2 was released a couple of weeks ago. I'm still trying to determine if it's worth the upgrade. All I can see that it does is increase performance and add some nice but not needed for me at the moment features. The move to 3.1 was worth it for me because there was a bug fix for the "code in an entry body gets run during blog update" bug but as far as I can tell there aren't any fixes of that level of importance in this latest release.

I'm probably going to hold on a version or two and see what happens. I have a hard time justifying the time to re-edit all the templates to put my customizations back just for the sake of a higher version number.

Don't get me wrong, if I was to start a new blog I'd use whatever was current. But, as it stands now, what I have works and works well for me. So, I guess I made made up my mind to not upgrade yet. I know there are other Nanoblogger users who read my blog from time to time. If you are one of them and you have upgraded can you comment on if it was worth the switch?

I'd been using Knoppix at work to bail out various Windows systems for the past few months so I decided to read Knoppix Hacks. I'd really only been using Knoppix to backup data before reinstalling Windows but after reading this book I have a few more tricks up my sleeve. My favorite ones and the ones I learned the most from are:

1. Wardrive with Knoppix
2. Check for Root Kits
3. Clone Hard Drives
4. Wipe a Hard Drive
5. Create Linux Software RAID

These are just a few of the 100 Hacks. I like the idea of checking for root kits from a read only file system and learned how to use dd for drive cloning. I've used partimage for cloning for a while but not every system will have that while every thing should have dd. And software RAID under Linux has always been sort of Voodoo for me but it turns out to be really easy. I know RAID isn't specific to Knoppix but at least I've finally learned something about it and tried it out on a system. The chapter on Repairing Linux is more useful and complete than the chapter on Repairing Windows but that really doesn't matter since the only effective way to repair Windows most of the time seems to be to back up data and reinstall the OS.

Now that I've done with this I'm not sure what I want to read next. I have a book on NIS and NFS that has been sitting on my shelf for some time waiting to be read, but I don't really have a pressing need to learn it at the moment. I'm sort of torn between learning something about SQL or sed and awk. But there are other things that I want to review and brush up on, like shell scripting in bash and getting better at using vim. I may buy the 3rd Edition of Learning bash since it covers bash 3.0 and read that to learn what's new and to review all at once.

After ages of using the same old K6-2 450MHz System that I built for him years ago my father finally gave in to the fact that he needed to upgrade his PC. Of course, this task came to me so I hopped on over to newegg.com and ordered a case, 512MB RAM, a AMD Semperon 2200+ CPU/Mobo combo set, a new CD-RW, and a 40GB Hard drive. The price was pretty good to, all that with shipping was under $300. Anyway, I tossed it all together and started installing his beloved Windows XP. And yes, I did try to get him to give Linux a shot. But as soon as I mentioned the "L" word he clammed up and didn't want to hear anything else about it. I'll never understand how closed minded some people are. But hey, he's the old-man and it is his computer. I just told him that when he gets tired of the spyware and viruses that I have something better for him. Well, at the least I've gotten him to stop using IE and run Firefox instead.

Anyway, here I am at the 2nd try to get Winblows Xpee to install on this brand new system. For some odd reason, immediately after a fresh install plus drivers, the system started acting strange. As in, it wouldn't let me install Service Pack 2 because "the cryptographic service isn't running" and when I attempted to actually view what services were running the computer management window would freeze up. The only thing it could possibly have been is the drivers, or maybe it's that Windows sucks? So, this time I guess I'll have to install Service Pack 2 right after the OS then install the drivers and see what happens. One of the drivers probably relies on fixes in SP2.

Yesterday I finished modifying stratdate and submitting the changes I made to the author. He seemed to like what I did plus I learned a few new things about scripting in bash. I don't know if/when he will put the changes into it officially, but I have the modified script in my downloads section in case anyone cares to give it a whirl. This is just a thing I did for the heck of it. I'd go to the projects page on sourceforge for the official one. I've upgraded a spare system as a test from a default Slackware 10.1 install to slackware-current w/o a hitch as a test, so far so good.

The changes I made are:

22/05/2005 - Brian Reichert (rignes@ptd.net)


Made copying of the Slackware 10.1 CD prior to running rsync work.
----
Put in a check for the kdei directory and will only move it out of
the $LPATH/slackware-current/slackware tree if it exists.
----
Added UPG_KERNEL to the conf file to choose whether to upgrade the
kernel or not.  I personally prefer to take care of my kernel
manually.  This also prevents the updating of alsa since the driver
modules are kernel specific.
----
Prevent the upgrading of aaa_elflibs* since Pat mentioned in the
Slackware 10.0 changelog that this is meant only for an initial
install and should never be updated
----
Added RUN_LILO in the conf file so you can choose to not run lilo
after each upgrade if you run GRUB or just prefer to do it yourself.

I've been spending a decent hunk of time watching what goes on, and occasionally partaking in, the goings on in ##slackware on freenode. Once again the talk of how evil upgrade tools are for slackware came up. Honestly, I'm good and sick and tired of hearing about it so I finally asked for examples on just what has caused this attitude. Well, it turns out that everyone who could tell me of any problem mentioned upgrading glibc while not in runlevel 1 has caused these problems. It is made clear in the UPGRADE.TXT on any Slackware CD that you must telinit 1 prior to upgrading. And all the tools that I know of don't check your run level and warn you of this. Now, I personally consider it a bad idea to blindly trust automatic tools, I admit it would be a nice check to have to prevent the moment when you have a brain fart and forget about changing your runlevel. As a side note, I think the reason I've had zero problems with swaret is because I always change runlevels for glibc.

There is one script called stratdate, which is based on Phrag's slackmon, that does just that. Basically, it gives you a nice tool to see what has changed in the slackware-current changelog since you last checked it, rsync it to your drive, and upgrade your system. It essentially does the exact same thing that UPGRADE.TXT explains. It looks nice, but there are a few things about it that aren't 100%. In particular, there is a feature that should copy the contents of Slackware CD1 to your drive prior to running rsync to help save download time that isn't working quite right. Also, as it is written now it will upgrade your kernels and automatically run lilo. This is nice if you run a 100% stock Slack, but what about if you run grub, or the 2.6.x kernel instead?

I'm currently tooling around with it as sort of an exercise in bash and will try to add the ability to skip a kernel upgrade and the running of lilo.

I keep up with slackware-current on my main system here, mostly because I like to see whats coming up. I don't mind because I keep good backups and slackware-current has been more stable then the supposedly stable releases of some other distributions I've tried out. I'm no programmer, so my knowledge of these things are limited, but the recent update to glibc to add Native POSIX Thread Libraries is supposed to offer significant performance increases in highly loaded systems. It's also gets rid of that annoying message that cedega keeps giving me about lack of pthreads support.

Anyway, I did this update, after a full system backup because I've heard horror stories about upgrading glibc and only ran into one hitch so far. It seems that xmms doesn't like the new libs. I tried compiling it myself and it still didn't like them. So I was forced to run "LD_ASSUME_KERNEL=2.4.30 xmms" to get it to run correctly. Furthermore, this problem only cropped up when running the nvidia drivers from nvidia.com. So, I figured I'd shoot an E-mail to Patrick Volkerding and let him know. Well, I was sure surprised when I got a reply from him about 10 minutes later with a link to a package he was working on to fix it. How's that for service eh? The thing is he knew what the problem was, but didn't know anyone who had nvidia cards to test out his fix and asked if I could. I tested it out on my spare system and it worked like a charm.

I know it's a minor thing, but it feels nice helping out just a little for the distro of my choosing. Anyway, I'll wait until the change makes it into the current tree before I use it on the main system. There seems to be quite a bit of action in -current the past few days. I like that!

Of course, all the doomsayers on ##slackware are still up in arms about any upgrade tool for slackware. And, once again, I've updated to current using swaret plus a little common sense with out a problem. I'm still waiting for it to kill my system like they say it's going to.

This morning I saw on slashdot a story about an ipod explosion. A short conversation about this on the LVLUG IRC got me a bit fired up and critical. So, here are my critical thoughts.

The title, "Teenager's ipod goes boom", brings pictures of well, things going boom. As in an explosion. So, I thought, "Wow, I never heard of such a small device causing an explosion! I better read this." After reading it I was disappointed. Sensationalism was used to suck me in and get my interest, then once I got the real facts I realized I just waisted time on something utterly useless and non-informative. Granted, it was only 5 minutes or so but still, that's not the point.

So, lets look at this a little. Here is part of the article:

Don't put your ipod through the washing machine. And if for some reason you do, don't try to fix it with a screwdriver.

That's the advice of fire investigators probing a small explosion that burned a hole in the bed of a Melbourne teenager who tried to perform emergency surgery on his ailing mp3 player.

OK, the important part is, "...probing a small explosion that burned a hole in the bed...". When I read this sentence I pictured shards of plastic flying and a hole in the bed big enough to put your head into.

Then it goes on to say:

"It wasn't working, the young fella tried to undo it or fix it with a screwdriver and at that stage there was an explosion, or more of a pop.

"It was more smoke than fire but it did leave a burn mark on the cover."

So, it wasn't really an explosion per se, but just a pop. And now the hole in the bed is actually nothing more than a burn mark on the cover. How stupid do they think people are? Do people actually not notice this kind of contradiction within the same article?

I finally finished HTML & XHTML: The Definitive Guide this weekend after having it shows as being currently read for months. I used to run a website back when HTML 1.0 was current and wanted to get my skills up to snuff just enough to keep my blog here nice and I can say this book has been very helpful.

HTML & XHTML is written such that it can be read either cover to cover or used as a reference. I read it straight through and noticed that each section has a fair amount of repeated info such as the lang and id tags. This is a good thing for a reference, but sort of annoying when just reading it like a regular book. Even so, once I felt I knew a particular repeated subject I'd just skip it, and if I didn't get it enough I'd read it again. They say repetition is the key to learning right?

All depreciated tags and browser specific extensions are marked clearly. I found myself skimming over the browser extensions just so I knew what they were but not really trying to learn them, mostly because I'm trying to stay away from them and stick with the XHTML 1.0 transitional standard. For tags or attributes that are depreciated it's made clear that they still work for now, but will probably not be usable forever and often suggests that CSS can be used to accomplish similar results.

Now, the most recent edition (5th) was first printed in 2002. and only talks about Netscape and Internet Explorer, which makes sense as they were the two big browsers at that time as I recall. While working through the examples I often saw comments that certain things were no longer supported in current browsers or not implemented yet. I tried a handful of the supposedly unsupported examples in Firefox and they worked perfectly. The point being that, even if the book says it's won't work that you should just try it and see. Even so, the unsupported tags are covered just as well as the supported ones for future reference. This was a good move I think because the core material won't be outdated nearly as fast.

One thing sort of puzzled me a little. The authors seem to play up XHTML 1.0 as being terribly complex compared to HTML 4.01. The main differences are that in XHTML every tag must be lowercase, have an end tag, and all attributes need to be enclosed in quotes. To me this isn't that harsh. Especially since when I was using HTML 1.0 I always thought the end tags were required. After drilling in this false complexity into your brain they then go on to suggest you use a good web authoring utility to do your web pages. This is said sometime after they show the output of a web authoring program compared to an identical handwritten page. The hand written one was about 20 lines or so, whereas the machine generated one was over a page long in the book. Then they say how important it is to go through your machine generated documents and clean them up. I think that if you need to know the tags well enough to be able to clean up a document then you should know them well enough to just skip the extra step and write your code nice by hand from the get go. There are valitator tools available from the W3C to check your work.

Two other things I like are the Tips and Tricks chapter which has some cute ways to use tables that I never thought of before and a nice pull out cheat sheet for people like me with no memory. Overall I'm happy I bought HTML & and XHTML: The Definitive Guide.

I know this has been all over the news but supposedly Microsoft wants to make peace with the Open Source community. I'm neither an analyst nor a lawyer, but my gut reaction is, "Yeah right." I mean, Microsoft is known for its embrace, extend and extinguish tactics. They've been this way from their beginning, which is something on the order of 20 years and I don't think they will ever really change at heart. Just like any other giant company they change only when they absolutely have to change as mandated by market or law. The only willing change they seem to make is when it benefits them in some way. That's called greed. They never change because it's the right thing to do, or because it would be better for the world, community, environment, etc.

My two cents is that this could be the beginning of the embrace phase and should be viewed with much caution. But that's just me.

Update: It look like there the author of this article at yahoo news feels the same way.

<rant>

I can sure feel for this fellow. This is exactly what I go through I'd say at least twice a month at work.

At the least the guy he worked with seems to have known what files to backup. The majority of the people who bring us computers have no clue where the hell they just saved their files. Usually I ask them about My Documents and they ask me what that is. It's an icon on the freakin' desktop! How can they miss it! The best fun with this is the files could be anywhere. I've seen systems where some files were saved in C:\Program Files\Microsoft Office, some On the desktop, and any of other places, usually not in My Documents where they are supposed to go. And, incidental, M$ Office is supposed to default to for it's save location.

Another favorite:

Brian: So, before I wipe your system, can you tell me what files you'd like backed up?

1user: What do you mean what files?

Yes, I've actually been asked that. Then they say I can't lose any files or their lives will be over.

If they were so darn important you'd think they'd know what they are, and where they are located. Oh, and maybe do something original and actually back the things up. *sigh*

</rant>

Sorry, just had to rant that out of my system. I feel much better now.

I think that if I had to pick one Firefox extension that was the most useful to me it would be Adblock. I use to both at work and at home. One thing that does annoy me about it though is that it doesn't come with any good filters that I know of. Actually, it doesn't come with any filters at all. Which got me Googling on various phrases to find a list. I could have generated a list on my own, but that would have taken time, and skillful use of regular expressions among other skills that I don't have. Besides, I wanted to block the ads NOW!

I found a few that were simply OK and I ended up tweaking them but still didn't have the results I wanted. I was either blocking stuff that I didn't want blocked, or ads were getting through and annoying me. That is until I stumbled on these Adblock Filters. The site isn't pretty, just a list of files, but the file -instructions.txt is very clear and the filters are updated oven, about once a week. One thing I almost missed though was the settings.txt file at the bottom of the list with suggested settings for Adblock itself.

I've been using these filters for close to a month now and I'm very pleased with the results.

Between this filter list, Adblock, and the popupsdie extension I can't say when I last noticed an ad of any sort while browsing.

In nanoblogger there is a really nice option to run HTML Tidy every time you add an entry. It's purpose is to go through your code, check for obvious errors and hopefully correct them. One of the things that nanoblogger does is put the XHTML Validated icon in if you want to show others that you are standards compliant. Well, while messing around I decided to actually click my own XHTML Validated icon and see what happened. Much to my surprise I wasn't even close to being valid. The validator found something on the order of 50 errors just on the front page! My first though was, "WTF? I'm running tidy! It should be OK."

My first reaction was to let the guys on the nanoblogger mailing list know that I was having a problem. That was my first mistake. I committed the sin of not investigating farther into the situation before I opened my big mouth. After a couple of posts I realized that the problem was indeed mine and no one elses. So, off I went to edit my page and see what I did wrong.

Now, there were plenty of errors on my part. Things like uppercase tags, ampersands that weren't encoded properly, missing end tags for things like images, and no alt attribute for the images either. However, as anyone can see, I've added things that are not part of the standard nanoblogger. In particular, I've added stuff from Amazon, Bloglines, and the Creative Commons License among a couple of other things. What I noticed is maybe about 75% of the errors I found were in code that either I didn't write or was never a part of nanoblogger in the first place. Amazon by far was the worst. When you get the code from them for their search box, or to put the images of books and things that I blog about from time to time, you just copy and paste it. I'm not talking just copying and pasting the address from the address bar on your browser. They actually have a page that generates the URL for you and put it in a text box for you to copy. You'd think that they would take the time to make it correct. But, apparently they do not. So, lesson learned, don't assume it's write. Take a few minutes and look it over.

As an interesting aside, I noticed that some of the tags were IE specific, thus not to the standard. I'm not sure why these tags are even there in the first place because when I took them out everything looked the same in both Firefox and IE.

So, I went on through all of the archives and ran the validator on them just to see what was what and, of course, found a slew of errors. Now, I know I could have just left all this go, but I mainly did it as a learning exercise. Also, I figure if I'm going to display those nice little graphics saying how wonderfully standards compliant I am then I probably should be.

So, anyway, lesson learned are:

• Don't blindly trust code you didn't write. At least look it over before you use it
• Don't point the finger at the tool, when it's the user (being me this time) who doesn't know how to use the tool correctly
• Investigate the problem before barking for help. It help you not look like an ass and you may just learn something in the process.

You'd think I'd know this all by now?

Well, now that I've learned something about CSS I've finally gone about making the photo gallery not suck. As usual, since I apparently have no personal creativity to speak of, I've taken heavy influence for it's appearance from my friend Ricardo's gallery. Incidentally, he also wrote jGal which I use to generate the albums themselves.

Hopefully I met my goal of making it not suck.

I've taken a moment to mess with the default style sheet for nanoblogger a little. I just finished the the CSS chapter in HTML & XHTML: The Definitive Guide and decided to put my new skill to some simple use.

I hope it doesn't look terrible. I admit, I did kind of borrowed the look of B10[m|g]'s Blog. I hope he doesn't mind.

Anyway, how does it look? Any suggestions?

I was just poking around on Slyck.com getting updated on the latest news about P2P networks when I found an article claiming to end 99% of illegal file trading.

Just on the little bit of info given in this article this seems like it has the potential, if done right, to be a short term hit on P2P networks. I say short term because it seems to take advantage of weak file hash algorithms in various P2P networks by creating a bogus file that happens to have the same hash as the file with the real, usable, content. I'm no expert by any stretch of the imagination when it comes to encryption and hash methods. But it seems to me that this could be rendered useless or at the very least much harder by simply using a different, well tested and trusted hash method.

I don't know what methods are available, but I see many uses of MD5 or SHA. Maybe one of these would do the job? Does anyone more knowledgeable than I on these matters read my blog? If so, feel free to comment, I'd like to know more on the subject.

Even so, this approach seems like it could do what it is intended to do if done right. That is, to harm the spread of illegal content while allowing legal content to flow. But, as with all things technology and P2P in particular, I'll just sit back and see what happens.